Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Jun 5.
Published in final edited form as: Cell. 2014 Jun 5;157(6):1380–1392. doi: 10.1016/j.cell.2014.05.009

Bromine is an essential trace element for assembly of collagen IV scaffolds in tissue development and architecture

A Scott McCall 1,†, Christopher F Cummings 2,3,†, Gautam Bhave 3,†, Roberto Vanacore 3,4, Andrea Page-McCaw 5,6,7, Billy G Hudson 2,3,4,7,8,9,*
PMCID: PMC4144415  NIHMSID: NIHMS602466  PMID: 24906154

Summary

Bromine is ubiquitously present in animals as ionic bromide (Br−) yet has no known essential function. Herein, we demonstrate that Br− is a required cofactor for peroxidasin-catalyzed formation of sulfilimine crosslinks, a post-translational modification essential for tissue development and architecture found within the collagen IV scaffold of basement membranes (BMs). Bromide, converted to hypobromous acid, forms a bromosulfonium-ion intermediate that energetically selects for sulfilimine formation. Dietary Br-deficiency is lethal in Drosophila while Br-replenishment restores viability, demonstrating its physiologic requirement. Importantly, Br-deficient flies phenocopy the developmental and BM defects observed in peroxidasin mutants and indicate a functional connection between Br−, collagen IV, and peroxidasin. We establish that Br− is required for sulfilimine formation within collagen IV, an event critical for BM assembly and tissue development. Thus, bromine is an essential trace element for all animals and its deficiency may be relevant to BM alterations observed in nutritional and smoking related disease.

Introduction

Basement membranes (BMs) are specialized extracellular matrices that are key mediators of signal transduction and mechanical support for epithelial cells (Daley and Yamada, 2013; Hynes, 2009; Lin and Bissell, 1993; Werb, 1997; Yurchenco, 2011). Indeed, BMs coordinate branching morphogenesis and define epithelial tissue architecture, facilitate tissue repair after injury, and guide pluripotent cells in tissue-engineered organ regeneration (Song and Ott, 2011; Vracko, 1974). Embedded within the BM is a sulfilimine crosslinked collagen IV scaffold that imparts functionality to the matrix of multicellular tissues in all animal phyla (Bhave et al., 2012; Fidler et al., 2014; Vanacore et al., 2009). Collagen IV scaffolds provide mechanical strength, serve as a ligand for integrins and other cell surface receptors, and interact with growth factors such as BMPs to establish signaling gradients (Wang et al., 2008). Mutations in the collagen IV scaffold cause BM destabilization and tissue dysfunction in humans, nematodes, flies, and mice (Borchiellini et al., 1996; Gould et al., 2005; Gupta et al., 1997; Hudson et al., 2003; Poschl et al., 2004).

Assembly of the collagen IV scaffold is an intricate processes of organization and covalent reinforcement. Triple-helical protomers extracellularly self-assemble into insoluble lattices, and nascent scaffolds are stabilized via the enzymatic formation of sulfilimine crosslinks between the NC1 domains of two juxtaposed protomers at residues methionine-93 (Met93) and hydroxylysine-211 (Hyl211)(Vanacore et al., 2009)(Fig. 1A). Peroxidasin, a heme peroxidase embedded within BMs, catalyzes the formation of sulfilimine crosslinks which confer critical structural reinforcement to collagen IV scaffolds, as seen in nematodes, flies, and zebrafish, where loss of peroxidasin causes BM dysfunction (Bhave et al., 2012; Fidler et al., 2014, Gotenstein et al., 2010).

Fig 1. Measurement of sulfilimine crosslink content within NC1 domains of collagen IV scaffolds.

Fig 1

Open in a new tab

(A) Diagram of the collagen IV scaffold, showing the relationship of NC1 hexamers sulfilimine crosslinks, peroxidasin (PXDN), and hypohalous acids (HOX). Inset shows resolution of dimeric (D1 and D2) and monomeric (M) NC1 domains by SDS-PAGE. Representative NC1 domains are shown from bovine placental BM (PBM), bovine glomerular BM (GBM), and murine collagen IV matrix produced in PFHR-9 cell culture

(B) High-resolution mass spectrum depicting the multiple oxidation states of tryptic peptides containing the sulfilmine (S=N) crosslink

(C) Extracted ion current (XIC) based quantitation of S=N crosslinked peptides from D1 and D2. Full data appears in Figure S1.

(D) Diagram showing the crosslinking status of observed NC1 banding in SDS-PAGE, where D1 is singly crosslinked and D2 is doubly crosslinked with a resultant higher electrophoretic mobility.

Peroxidasin forms hypobromous acid (HOBr) and hypochlorous acid (HOCl) from bromide and chloride, respectively, both of which can mediate crosslink formation (Fig 1A). In vitro studies point to a preference for Br− during enzymatic sulfilimine formation but its role within the in vivo reaction is unknown, particularly in light of the vast excess of Cl− over Br− in most animals (Weiss et al., 1986). Despite its ubiquitous yet trace presence within animals, Br− is without a known essential function. Bromide is a cofactor for eosinophil peroxidase (EPO) following eosinophil activation (Mayeno et al., 1989; Weiss et al., 1986), but the relevance of this is unclear as EPO preferentially oxidizes SCN− over Br−.(Nagy et al., 2006). Thus, the definitive identification of Br− as a cofactor for peroxidasin-mediated crosslink formation would represent the first essential function for the element bromine.

Herein, we generated Br-free Cl− salts and found that peroxidasin uses Br− to catalyze formation of sulfilimine crosslinks with at least 50,000-fold greater efficiency compared to Cl−. Drosophila raised on Br− deficient diets resemble peroxidasin loss-of-function mutations, including developmental abnormalities, lethality, and altered BM and tissue morphologies. Importantly, replenishment of Br− to the diet restores the normal phenotype in all metrics addressed here. Collectively, our findings establish a physiologic requirement for Br− in animals, the mechanism by which Br− functions, and the essentiality of a sulfilimine-crosslinked collagen IV scaffold in the assembly and function of BMs.

Results

Structural basis for sulfilimine crosslink heterogeneity in the collagen IV scaffold

Sulfilimine crosslinks join NC1 domains at the interface of two adjoining triple helical protomers within the collagen IV scaffold, forming a globular hexameric structure. Dimerized NC1 domains may be bound together by either one or two sulfilimine crosslinks, due to the presence of Met93 and Hyl211 in both domains. Following biochemical isolation and SDS-PAGE analysis, NC1 hexamers dissociate into crosslinked dimers, termed D1 and D2, and uncrosslinked monomers (Fig 1A). The structural distinction between D1 and D2 is unknown but of longstanding interest (Langeveld et al., 1988). Since HOBr reacts with uncrosslinked NC1 hexamer to uniquely form D2 with a pattern resembling native placental and glomerular BMs, we endeavored to structurally define the D1 and D2 isoforms of crosslinked hexamer. We hypothesized that D1 and D2 differed by the number of crosslinks, D1 having one and D2 with two crosslinks. Using liquid chromatography-mass spectrometry (LC-MS) to determine the abundance of crosslinks in D1 and D2, we found 1.95-times greater sulfilimine-containing peptides in D2 relative to D1 (Fig. 1B–C, S1), indicating that D2 has two and D1 has one crosslink. We thus used the relative abundance of D1 and D2 on SDS-PAGE analysis of NC1 hexamers to assess sulfilimine crosslink content (Fig 1D).

Bromide is required for sulfilimine formation

We next examined the effect of halides (F−, Cl−, Br−, I−) and the pseudohalide thiocyanate (SCN−) on sulfilimine crosslinking in cell culture. SCN− and I− inhibited the reaction while Br− enhanced crosslink formation (Fig 2A and S2). Since F− proved cytotoxic, and background levels in media precluded testing of Cl−, we moved to isolated PFHR-9 BM as an in vitro model for crosslink formation. In this model, KI inhibits peroxidasin to generate an uncrosslinked scaffold, and subsequent KI removal with addition of H2O2 drives crosslink formation by peroxidasin. (Bhave et al., 2012). F− was inert to crosslink formation, so we used 100 mM KF as an ionic control while titrating either Cl− or Br− (Fig 2B). Br− robustly catalyzed crosslink formation at 10 µM while Cl − remained inactive until 100 mM (Fig 2B).

Fig 2. Bromide is the required cofactor for sulfilimine crosslink formation.

Fig 2

Open in a new tab

(A) The effect of halide ions on sulfilimine crosslink formation is examined in PFHR-9 matrix. Inhibition values were calculated from non-linear curve fitting: KI (IC50= 84µM 95%CI[30–241µM]), KSCN (IC50=17µM 95%CI[3-24µM]). Contrasting with these effects, exogenous potassium bromide (KBr) enhanced the reaction. Points represent mean ± S.D. (n=3). See also Fig. S2.

(B) Uncrosslinked PFHR9 matrix was crosslinked in vitro in the presence of KCl and KBr. Reacted for 1 hr. at 37°C with 1 mM H2O2 100 mM KF used as ionic strength control. Collagenase digest analyzed by SDS-PAGE and Coosmassie staining.

(C) Schematic of Br-free Cl- salt purification apparatus and setup. Resulting salt was analyzed by ICP-MS for bromide content. Further analysis of salt reagents appears in Table S1 and Figure S3

(D) Crosslink formation in PFHR-9 matrix with Br-free KCl. Reaction buffer contained 10 mM phosphate buffer (pH 7.4), 100 mM Br-free or reagent grade KCl, and 1 mM H2O2 and 200 µM PHG where appropriate. Displayed SDS-PAGE gels were stained with Coomassie blue.

(E) Sulfilimine (S=N) crosslink formation in PFHR-9 cell-culture tested under Br-free conditions. Culture conditions and media formulations are presented in detail in the supplementary methods. NC1 hexamers were isolated via collagenase treatment and analyzed by SDS-PAGE. The amount of crosslinks per hexamer is graphed as the mean±95% C.I. (n=3). All sample groups had equal variance, one-way ANOVA was performed (p<0.001) and differences between groups was tested using Tukeys post-hoc analyisis (***p<0.001)

While these data pointed to a strong preference for Br− over Cl− in crosslink formation, we considered whether contaminating Br− in our Cl− solutions might confound the results. Indeed, we measured Br− content at 5.91 µM/100 mM KCl (ICP-MS)(Table S1), making the apparent Cl− activity difficult to distinguish from that of contaminating Br−. To address this, we produced Br-free NaCl and KCl (<11.4 nM Br−) (Fig 2C, Fig S3A, Table S1). Intriguingly, Br-free Cl− did not support crosslink formation (Fig. 2D; Fig. S3B), while addition of 5 µM Br− rescued crosslink formation.

For validation, we tested the efficiency of crosslink formation in PFHR-9 cells grown in Br-free media (<2.5 µM Br−, as measured by Neutron Activation Analysis (NAA), Table S2–4). In Br-free and Br-added culture conditions, there was no appreciable difference in cell proliferation, cell viability, or collagen IV production. Importantly, Br-free media did not support formation of crosslinks in the collagen IV matrix yet the addition of 100 µM Br− to the same media rescued normal crosslink formation (Fig. 2E), establishing a requirement for Br− in sulfilimine formation.

Peroxidasin catalyzes sulfilimine crosslink formation via bromide

Continuing our studies with purified Cl−, we examined the sulfilimine formation capability of peroxidasin, myeloperoxidase (MPO), and EPO in Br-free conditions. Generally, peroxidases use peroxide to oxidize a halide ion to the corresponding hypohalous acid (HOX, X=Cl or Br) with MPO preferentially oxidizing Cl− to HOCl and peroxidasin and EPO forming HOBr via oxidation of Br−. After normalization of peroxidase activity of all enzyme preparations (Bozeman et al., 1990), we found peroxidasin to be much more effective in forming sulfilimine crosslinks within NC1 hexamers than MPO or EPO under identical conditions, especially regarding D2 (two crosslinks) formation (Fig 3A). Crosslinking by peroxidasin exceeded EPO despite normalized enzyme activity.

Fig 3. Peroxidasin uses physiologic Br levels to form sulfilimine crosslinks.

Fig 3

Open in a new tab

(A–D) Mammalian peroxidases are compared for ability to crosslink collagen IV NC1 domains in Br-free 1x PBS, (recombinant human peroxidasin (hPXDN), myeloperoxidase (MPO), eosinophilperoxidase (EPO).Uncrosslinked NC1 domains were isolated from PHG-treated PFHR-9 cultures. All peroxidase activity enzymes was nominalized by TMB assay prior to assay (Bozeman et al., 1990). Reactions proceeded for 10 minutes at 37° C after initiation by the addition of H2O2 and quenched with 5 mM PHG, 0.2 mg/ml bovine catalase, and 10 mM methionine. Gel representative of 3 experiments.

(A) Coomassie stained gel of enzymatic crosslink formation under reagent grade and Br-free conditions.

(B) Quantitative analysis of crosslinks formed per NC1 hexamer by MPO and hPXDN under Br-free and Br-added (100 µM) conditions. Data shown as mean ± 1 S.D. n=2. Student’s T-Test was performed (n=2) due to equal variance between groups.

(C) Effect of bromide titration on the proportion of D1 (1 crosslink) and D2 (2 crosslinks) NC1 populations following reaction with MPO and hPXDN. For reference, the proportions of D2 found in PBM and GBM are denoted on the graph. Data shown as mean ± 1 S.D. (n=2).

(D) Crosslinking efficacy of peroxidasin measured as crosslink formed per hexamer upon Br titration. EC50 value ± 95% C.I. (n=2).

We sought to characterize the responsiveness of peroxidasin and MPO to Br− levels. In Br-free saline, where HOCl is the only hypohalous product of either enzyme, only minor amounts of crosslink were produced by the enzymes. The addition of 100 µM Br− significantly enhanced crosslink formation by peroxidasin, as did increased H2O2 levels, (Fig 3B), suggesting that the in vivo enzymatic mechanism is responsive to both Br− and oxidant concentrations. Upon Br− titration, D1 formed prior to D2 indicating a sequential crosslinking mechanism (Fig 3C), complementing our LC-MS studies that revealed one and two sulfilimine crosslinks in D1 and D2, respectively (Fig. 1B – C). Thus, the relative amount of D2 represents a key index of crosslinks within the overall NC1 hexamer, and is a notable feature of tissue-isolated collagen IV (Fig. S4; [Langeveld et al., 1988]). In vitro, we found that crosslink formation by either a Br-H2O2-peroxidasin system or the direct HOBr application produces D2 to a similar degree as observed in tissues (Fig. S4).

Considering the trace levels of Br− in physiology, we calculated the EC50 for Br− in this system to be 4.5 µM (95% CI 3.8–5.2 µM) in the presence of 140 mM Cl− (Fig. 3D). Using MPO as a baseline for the efficacy of Cl-based oxidants, these data indicate a >50,000-fold efficacy difference for Br− over Cl− as a cofactor in the peroxidasin-catalyzed formation of crosslinks, demonstrating selectivity for Br−. Within the normal serum range of 10–100 µM Br− (van Leeuwen and Sangster, 1987), peroxidasin formed crosslinks at more than 90% of the available sites but was markedly less effective below this range, indicating an optimal Br− level for in vivo crosslink formation.

The chemical basis for selection of Br− over Cl− as the cofactor in crosslink formation

To investigate the chemical basis for the selectivity of Br− over Cl− in this reaction, we used the halogen-based synthesis of dehydromethionine, a cyclic sulfilimine product, as a mechanistic framework wherein a methionine halosulfonium intermediate (HSI) reacts either with an amine or water to form a sulfilimine bond or sulfoxide respectively (Fig. 4A , Armesto et al., 2000; Peskin et al., 2009; Young and Hsieh, 1978). We hypothesized that 1) collagen IV sulfilimine bond formation proceeds via an HSI at Met93 and that 2) selectivity resides with a bromosulfonium intermediate which predominately reacts with the ∊ -NH2 of Hyl211 to form the crosslink, whereas the chlorosulfonium intermediate predominantly reacts with water to form methionine sulfoxide, precluding crosslink formation.

Fig 4. Chemical mechanism of sulfilimine (S=N) formation within the NC1 hexamer.

Fig 4

Open in a new tab

(A) Working model of the oxidative formation of either sulfilimine crosslinks or methionine sulfoxide. kS=O and kS=N refer to rate constants in the formation of sulfoxides and sulfilimines, respectively.

(B) Uncrosslinked NC1 hexamers (5 µM) were reacted with hypohalous acids for 5 minutes at 37˚ C and the products analyzed by SDS-PAGE. Values represent mean ± 95%CI. (n=3).

(C) Uncrosslinked NC1 hexamer (1.3 µM) was reacted with indicated amounts of HOCl for 1 minute at 37° C in Br-free 1x PBS, followed by subsequent treatment with of 8 mol eq. HOBr (or HOCl as a control) and reacted for an additional minute at 37° C. Reactions were quenched with 20 mM methionine. Gel is representative of two experiments.

(D) The sequential model for D1 and D2 formation within the NC1 hexamer following complete stoichiometric oxidation of Met93. P1-P4 indicate the proportional probabilities of forming the observed products. Calculations are presented in Extended Experimental Methods.

(E) Free energy landscape for S=N formation within the NC1 hexamer based on the model outlined in panel D and Fig S5–6.

(F) Outline of overall chemical pathway governing the intrinsic chemical reactivity of S-Br and S-Cl at Met93.

In vitro, HOBr effectively promoted crosslink formation in a dose-dependent manner while HOCl poorly formed NC1 crosslinks (Fig 4B). We used mass spectrometry to determine the oxidation state of Met93 within HOCl-reacted monomers and found increased amounts of methionine sulfoxide (Fig. S5). To test whether methionine sulfoxide is indeed a “dead end” with respect to crosslink formation, we treated uncrosslinked NC1 hexamers with HOCl prior to HOBr treatment and found a dose-dependent inhibition of crosslink formation until it resembled treatment with HOCl alone (Fig 4C). Thus, both HOBr and HOCl target Met93 yet the latter oxidant creates an uncrosslinkable product, namely methionine sulfoxide. As further validation, similar results were obtained when sulfoxide was generated using prolonged treatment with high concentrations of H2O2 (Fig. S5G).

Probing deeper into the mechanism, we sought to characterize the reactivity and energetics of Met93 after oxidation with HOBr or HOCl. We modeled distinct HSI reaction pathways yielding sulfoxide or sulfilimine and two potential crosslinking events between opposing NC1 subunits (Fig 4D and Fig S6). Using densitometric analysis of SDS-PAGE gels, we measured the relative proportion of uncrosslinked, singly crosslinked, and doubly crosslinked NC1 subunits following complete oxidation with either HOBr or HOCl to calculate the sulfilimine and sulfoxide product ratios for both oxidative events (Fig S6C–D). The data indicated that the two oxidation events are not independent (Table S5) but, rather, formation of the first sulfilimine enhances the probability of a second crosslinking event. Possible physical interpretations of these data are that the first crosslink imposes steric constraints on the orientation of, Met93 and Hyl211 at the second site, or simply increases their spatial proximity such that the apparent local amine concentration increases 3 to 7-fold (see Extended Experimental Methods). We examined the relative free-energy difference in the transition states for the competing sulfilimine and sulfoxide reaction pathways (Figure 4E; Extended Experimental Methods; Figs S6B–D)(Seeman, 1983), and found that the bromosulfonium (S-Br) intermediate encountered a lower energetic barrier to sulfilimine formation for both oxidative events while the chlorosulfonium (S-Cl) intermediate faced an unfavorable barrier to sulfilimine formation relative to methionine sulfoxide (Fig 4E, Fig S6D).

Further explanation for the difference in products between Br− and Cl− in collagen IV may be found in the distinct chemical behaviors of S-Br and S-Cl. Experimental and in silico studies indicate that S-Cl species have highly polar transition states, and therefore participate in charge-controlled reactions which prefer ‘harder’ nucleophiles such as H2O relative to amine, and thereby favor sulfoxide formation. Conversely, S-Br species generate transition states with smaller partial charge, which favor orbital-controlled reactions that select for ‘softer’ nucleophiles (here understood as NH2-R relative to H2O) and thus prefer sulfilimine formation (Chmutova et al., 1999; Klopman, 1968; Pearson, 1968). Taken together, sulfilimine formation is thermodynamically favorable via the selectivity of an S-Br intermediate at Met93, providing chemical basis for the Br− requirement.

Bromide is essential for Drosophila development

Based on the chemical requirement for Br− in collagen IV sulfilimine bond formation and the conservation of the crosslink in multicellular tissues (Fidler et al, 2014), we hypothesized that Br− is essential for stabilizing tissues. We tested this hypothesis in Drosophila. Because standard Drosophila media contains ~15 µM Br−, we prepared a custom diet (Table S7) in which final dietary Br− was undetectable by NAA (Table S2). To address the impact of Br− deficiency over multiple generations, we raised flies on Br− free media, and compared their development to cohorts raised on either similar media with Br− supplementation or standard media (Fig. 5A). Initial maternal Br− contribution in embryos was 24.3 µM (Table S2) on the standard diet. After moving embryos to the indicated media, Generation 1 larvae grown on Br-free conditions exhibited developmental delay (Fig. 5B), yet development rates were similar between Br-added and standard media. Adult Generation 1 flies that survived were maintained on the same diet for 14 days to continue Br− depletion, and progeny Generation 2 larvae showed significantly reduced survival in Br-free versus standard; the phenotype was rescued in Br-added diet (Fig. 5C). Thus, Br− is essential for development in Drosophila.

Fig 5. Bromide is essential for development and basement membrane architecture in Drosophila.

Fig 5

Open in a new tab

(A) Generational Br-depletion scheme

(B) Generation 1 survival and time-to-development curves for w1118 flies on the standard diet vs experimental diets. Embryos from mothers fed a standard diet were placed on the indicated diet, and progeny were scored every 24 hours. There was not a signifigant difference in survival between groups by Log-Rank test (Left panel). The Br-free+ 100µM Br diet supported the same timing of development as the standard diet, whereas the Br-free diet caused a significant delay (p<0.001 compared to both standard diet and Br-free +100 µM Br) prior to pupariation (8 days) and eclosion (14 days) (Right panel). Data plotted as the group median ± interquartile range. N=30 for each group. 2 way ANOVA test showed a significant difference for pupariation and eclosion (p<0.001 ) §=different from standard, ‡=different from Br-free+100µM Br−.

(C) Generation 2 developmental survival on experimental diets. n>40 for each cohort. Tested by Log-Rank test.

(D) Percentage of eggs (mean +/− 95% C.I.) completing embryogenesis from mothers reared on Br-freeDEP or Br-freeDEP +100 µM Br diets for 5 days. In the Br-freeDEP experimental group, mothers were fed Br-free synthetic diet containing 80 mM total NaCl (Br-freeDEP) for 3 days prior to egg collection. The Br-freeDEP+100µM Br was treated in the same manner except that 100 µM NaBr was added to all food components of the Br-freeDEP synthetic diet. Hatching rate differences were observed for eggs collected 3–7 days after maternal diet implementation. n=300 eggs. Analyzed by the Mann-Whitney U test.

(E) Survival curve for w1118 flies under Standard, Br-freeDEP , and Br-freeDEP+100µM Br− dietary conditions. The survival difference between groups was highly significant (log rank test, n>40 for each group).

(F) Western blot of isolated NC1 domain from larvae treated as in (E), probed with an anti-Drosophila NC1 polyclonal antibody (Extended Experimental Methods). Associated larval Br-content was measured by EINAA (additional data in Table S2). Bonds/hexamer were calculated from the Western blot.

(G–I) Representative images of vkg454-GFP homozygous larvae reared under the conditions tested in (E) demonstrating holes in the BM (indicated by orange arrows) in the distal posterior midgut of Br-freeDEP larvae. Optical sections of mid-lateral gut plane visualizing the circular muscles in cross-section (f-actin stained with phalloidin) surrounded by a collagen IV (Vkg454 -GFP) scaffold and the gut epithelial BM. Gut lumen is oriented at the top of the image, anterior-posterior axis is horizontal. *= BM defect. Whole gut images, scale bar = 20 µ, mid-lateral plane optical sections, scale bar = 10 µ.

(J–M) Representative images of posterior midgut of 4 day old larvae with indicated peroxidasin genotype on standard food. Collagen IV anti-NC1 staining on non-permeablized samples, demonstrating similar BM staining as Vkg454-GFP (J). F-actin was stained with phalliodin. No muscle actin staining is visible in (M) due to muscle death, directly visualized by EM in (R). *= BM defect. Whole gut images, scale bar = 20 µ, mid-lateral plane optical sections, scale bar = 10 µ.

(N–R) Electron micrographs of circular sections through the posterior midgut, focusing on the BM (magenta psuedocolor) beneath the enterocyte (En) near a longitudinal muscle belly (LM). Trachioles (Tr) are occasionally visualized. Standard diet control (N) has a compact, normal BM. BMs are thickened and irregular in Br-freeDEP (O), PxnMI01492(Q), and Pxnf07229 (R); BM is similar to control in Br-freeDEP+100µM Br− (P). BMs from 15 independent sections for each group were evaluated for thickness and the histograms plotted. Scale bar = 0.5 µ.

(S) Distribution curves and parameters from fitted distributions of each experimental group. Optimal Box-Cox transformations were performed for normality and the transformed distributions were fitted. The curves to fit the data were significantly different by the F test. Pairwise comparison revealed no significant difference between standard and Br-freeDEP+100µM Br− while both curves differed significantly from Br-freeDEP, PxnMI01492, and Pxnf07229. Bootstrapping was performed to obtain the 95% C.I.s for the mean and S.D. for each group, revealing that the variance in BM thickness in Br-freeDEP, PxnMI01492, and Pxnf07229 is substantially higher. (See also Fig. S7)

Seeking to accelerate Br-depletion, we fed flies a Br-free diet containing elevated NaCl levels to reduce Br− half-life in vivo via halide flux seen in mammals (Pavelka et al., 2005). Female Drosophila were placed on a Br− depleting (Br-freeDEP) diet with or without supplemental 100 µM Br− prior to egg deposition, and the dietary conditions were maintained throughout progeny development. Initially, the Br-freeDEP egg cohort had a significantly reduced hatching percentage relative to Br-freeDEP+100µM Br− (Fig. 5D), suggesting that Br is required for successful embryogenesis. Nearly all hatched larvae died prior to eclosion under Br-freeDEP conditions, while 100 µM Br− rescued development to adulthood (Fig 5E). NAA analysis confirmed lower Br− levels in third instar larvae (3.4 vs. 23.6 µM for controls) in Br-freeDEP conditions (Fig 5F).

In Br-freeDEP conditions, we assessed the impact of Br-deficiency on crosslink formation in vivo. We used vkg454-GFP flies in which the single collagen IV α2 gene locus contains a GFP insertion near the 7S domain. We grew these vkg454-GFP flies on Br-freeDEP, Br-freeDEP+100µM Br− and standard diets and biochemically assayed sulfilamine-bond content via immunoblot (Extended Experimental Methods). We found grossly reduced sulfilimine-bond content in the Br-freeDEP cohort, which was rescued with Br− supplementation (Fig 5F, Fig S7A). Thus, Br− promotes sulfilimine formation in vivo.

Peroxidasin (Pxn) mutants have reduced amounts of collagen IV sulfilimine crosslinks and consequent perturbation of midgut BM, as shown previously (Pxnf07229, Bhave et al, 2012). Predicting that Br-depletion would phenocopy the Pxn mutant, we compared the midgut from Br-freeDEP larvae to two independent mutants of Pxn (PxnMI01492 and Pxnf07229). While normal scaffold architecture was seen on standard diet (Fig. 5G), Br-freeDEP conditions displayed gross disruptions (Fig 5H, red arrows) and splitting in the overall collagen IV scaffold (Fig. 5H, asterisk). Both phenotypes were rescued in Br-freeDEP + 100µM Br− media (Fig. 5I). Significantly, similar disruptions were seen in Pxn mutants using an anti-NC1 antibody (Fig. 5J – M). The Pxnf07229 phenotype appeared more severe than PxnMI01492, and the PxnMI01492 /Pxnf07229 transheterozygote demonstrated an intermediate phenotype. These data indicate similarities between Br-deficiency and the loss of peroxidasin.

We used transmission electron microscopy (TEM) to compare the BM ultrastructure in Br-depleted larvae with Pxn mutant larvae. Larvae raised on standard diet exhibited normal enterocyte and BM structure (Shanbhag, and Tripathi, 2009)(Fig. 5N). In Br-freeDEP larvae, the BM was irregular, thickened, occasionally diffuse, and wavy in various sections (Fig. 5O). Strikingly, both Pxn mutants exhibited irregular and thickened BM similar to the Br-freeDEP cohort (Fig 5Q – R). Moreover, Br-freeDEP conditions and Pxn mutants displayed circular muscles protruding into and deforming the gut epithelium, (Fig S7C,E,F), mirroring the actin staining in circular muscles (Fig 5H,K,M). All sections from the Br-freeDEP+100µM Br− and standard diet cohorts displayed normal BM and circular muscle morphologies (Fig 5N,P; Fig S7B,D). We quantified the BM morphologic changes observed by TEM, finding similar BM thickness in the standard and Br-freeDEP+100µM Br− diets but significantly thicker BMs in Br-freeDEP and both Pxn mutants (Fig 5S, Fig S7G).

Br-freeDEP conditions phenocopy the genetic loss of Pxn, so we hypothesized that Br− and Pxn interact in vivo to strengthen collagen IV scaffolds. It has been reported that collagen IV acts during Drosophila oogenesis as a “molecular corset” to control egg shape, restricting circumferential expansion so that egg growth promotes elongation along the anterior-posterior axis (Fig. 6A)(Haigo and Bilder, 2011). In eggs from mothers fed varying concentrations of Br−, we found a dose-dependent relationship between Br− and aspect ratio (Fig 6B) after approximately four days, consistent with a long biologic half-life for Br−. Interestingly, the aspect ratio of eggs on the Br-added diet (100 µM Br−) exceeded the ratio for eggs on standard diet (NAA measured 15µM Br−, Table S2)(Fig. 6B) suggesting that elevated Br− promotes additional sulfilamine formation to enhance tensile strength in the collagen IV molecular corset.

Fig 6. Br− and peroxidasin interact in vivo to strengthen collagen IV scaffolds.

Fig 6

Open in a new tab

(A) Schematic overview of polarized collagen IV scaffolds (molecular corset, green) which determines aspect ratio in Drosophila eggs.

(B) Br− concentration effect on egg aspect ratio, in single age-matched cohort of w1118 flies, over time. Vertical axis represents mean aspect ratio (±S.E.M). At 192 hours (inset), egg aspect ratio had increased proportionally to Br− concentration, with similar aspect ratios in 15 µM added NaBr and standard diet (measured as 15 µM Br− by NAA). Inset plotted as mean±95% C.I. and significance calculated with the Kruskall-Wallis test. Dotted line indicates egg aspect ratio reported by (Haigo and Bilder, 2011).

(C) An irreversible peroxidasin inhibitor, PHG, causes a dose-dependent reduction in the exaggerated egg elongation caused by excess dietary (100µM) Br. PHG was administered in the food. All wild-type (w1118) mothers were from the same cohort and reared identically, then divided into sub-cohorts for exposure to the indicated experimental diet. Significance among the conditions calculated using the Kruskal-Wallis test. Data plotted as mean±95% C.I. (image; scale bar = 500 µm). Dotted line indicates reported value for egg aspect ratio (Haigo and Bilder, 2011). All groups also differed significantly when compared individually using Dunn’s multiple Comparison testing (p<0.05).

(D–E) Pxn is required for Br-induced egg elongation. Two independent temperature-inducible RNAi constructs targeting Pxn were expressed in adult females fed 100 µM added Br−. Aspect ratios from maternally expressed PxnRNAi were significantly different than sibling-matched controls after induction for 72 hours (29° C) by Mann-Whitney U test. *p<0.05, ***p<0.001. Data plotted as mean as mean±95% C.I., (inset image scale bar = 500 µm). Dotted line indicates reported value for egg aspect ratio (Haigo and Bilder, 2011).

(F) Egg aspect ratio on standard diet and synthetic Br-freeDEP and Br-freeDEP + 100µM Br− diets. Eggs were collected after mothers were fed indicated diet for 7 days. Differences in egg aspect ratio were observed in eggs collected after 5–7 days of experimental diets. Representative pictures of eggs are shown (scale bar = 500 µm). Aspect ratio plotted as mean +/− 95% C.I. (graph; Mann Whitney U Test **p<0.01 ***p<0.001). Dotted line indicates reported value for egg aspect ratio (Haigo and Bilder, 2011).

(G) Collagen IV density appears normal in eggs from Br-depleted mothers. BM of stage 8 egg chambers from mothers expressing Vkg-GFP and fed the indicated diet (confocal images). For quantitation, fluorescence intensities of z-stack projections were summed in areas where the whole thickness of the BM had been observed, and normalized to the observational area. n=9 for each group, There was no difference in the medians between the groups by the Kruskal-Wallis test. Data plotted as mean ±95% C.I.., (image; scale bar = 20 µm). (See also Fig. S7)

We used this elongated egg aspect ratio to probe whether Br− and Pxn act via a common mechanism in strengthening collagen IV. We asked if Pxn is required for the elongation phenotype via two methods. First, we used an irreversible inhibitor, phloroglucinol, to inhibit peroxidasin activity, and we observed a dose-dependent suppression in egg aspect ratio in the presence of elevated Br− (Fig. 6C). Second, to confirm the specificity of this interaction, we used two separate ubiquitously driven temperature sensitive conditional RNAi constructs to knockdown peroxidasin in adult females in the presence of 100 µM Br−. In both RNAi experiments, aspect ratio was significantly decreased relative to controls 3–4 days after the onset of Pxn knockdown (Fig 6D–E, S7) while controls displayed normal augmentation of aspect ratio under identical conditions. Thus Pxn is required for the Br-induced elongation phenotype. To address the alternative hypothesis that Br− levels modulate collagen IV deposition, we examined Vkg-GFP immunofluorescence in eggs from mothers raised on Br-freeDEP media. Like the Br-deficient diet, the Br-freeDEP media reduced egg aspect ratio (Fig. 6F), but collagen IV content appeared similar to controls (Fig. 6G) after one week of maternal exposure to Br-freeDEP diet, suggesting that the egg aspect-ratio phenotypes are caused by structural deficiencies within the scaffold.

Discussion

Essentiality and Function of Bromide in Animals

We provide evidence that bromine is essential in animals, satisfying the principal requirements for elemental essentiality: 1) Element deficiency leads to physiologic dysfunction, 2) Repletion of the element reverses dysfunction, and 3) Biochemical explanation of the physiologic function (Mertz, 1981). Br-deficient Drosophila display altered BM and tissue morphology, aberrant embryogenesis, larval mid-gut defects, and lethality, while Br− repletion restored normal development. Mechanistically, the assembly of crosslinked collagen IV scaffolds requires Br−.

Sulfilimine crosslinked collagen IV scaffolds are central to the form and function of BMs in animals (Bhave et al., 2012; Fidler et al., 2014). Our data indicate that the crosslink stabilizes nascent collagen IV scaffolds, effectively modulating scaffold assembly and BM thickness. Since sulfilimine formation involves the concerted activity of collagen IV, Br−, peroxidasin, and oxidant, we view each as critical for BM assembly and tissue development (Fig 7A–B).

Fig 7. Model of the essentiality of bromine in forming collagen IV sulfilimine crosslinks.

Fig 7

Open in a new tab

(A) Diagrammatic relationship between collagen IV sulfilimine formation and tissue phenotype.

(B) Schematic representation of role of bromide in oxidative formation of sulfilimine crosslinks.

(C) Proposed chemical mechanism of sulfilimine formation by HOBr.

Mechanistic Role of Bromide in Sulfilimine Formation

The requirement for Br− during sulfilimine formation derives from the selectivity of the bromosulfonium reaction intermediate. The chemical character of bromine uniquely creates an energetically favorable reaction between the S-Br intermediate and Hyl211. The S-Br molecular orbital structure facilitates selective reactivity with an amine nucleophile to form the crosslink, contrasting with the highly polar S-Cl intermediate that preferentially forms a sulfoxide via charge-controlled reaction with water (Fig 7C). Peroxidasin harnesses this HOBr-based selectivity during crosslinking while apparently avoiding oxidative damage to the BM.

Bromide Homeostasis

Br− is mainly located extracellularly and has been used in the clinical measurement of extracellular volume (Barratt and Walser, 1969; Brodie et al., 1939). Plasma Br− is 67 µM in healthy people, congruent with Br− concentrations that support sulfilimine formation in flies, and are maintained within an order of magnitude in many species (freshwater fish [Woods et al., 1979], flies [Piedade-Guerreiro et al., 1987], rodents [Van Logten et al., 1974], and humans [Olszowy et al., 1998; van Leeuwen and Sangster, 1981]). In humans, plasma Br− is maintained via diet and renal excretion (Trautner and Wieth, 1968; van Leeuwen and Sangster, 1981; Walser and Rahill, 1966; Wolf and Eadie, 1950). Drosophila likely conserve Br−, possibly contributing to the timeline of phenotype development in our generational dietary Br-deficiency model (Fig. 5A–C). Dietary Br-deficiency has been suggested to suppress tissue growth and increase lethality in goats (Anke et al., 1990), while high serum Br− (>12 mM) causes neurologic and dermatologic complications (van Leeuwen and Sangster, 1987). Taken together, this implies that an optimal Br− concentration might exist and is regulated in vivo.

Clinical Implications of Bromide Deficiency

Bromide deficiency may have implications in human health and disease. Patients receiving total parenteral nutrition (TPN) are reported to have low plasma Br− levels due to lower dietary Br consumption (Dahlstrom et al., 1986), and end-stage renal disease patients have enhanced Br− losses as a consequence of dialysis (Miura et al., 2002; Oe et al., 1981; Wallaeys et al., 1986). Since Br has not been considered an essential trace element, systematic investigations on Br− replacement have not been pursued in these disease states. Intriguingly, TPN alters intestinal mucosal architecture and function in a manner reminiscent of the mid-gut phenotypes of Drosophila Pxn mutants and Br-deficient larvae (Groos et al., 2003). Furthermore, functional Br-deficiency may occur in smokers in spite of normal plasma Br− levels because of elevated levels of serum SCN−, which inhibits sulfilimine bond formation. In our present study, we found SCN− to be a potent inhibitor of peroxidasin-mediated crosslink formation (Fig.2A and S2). Therefore, in some smokers with elevated SCN− levels (130 µM, 1 pack per day) (Tsuge et al., 2000), reinforcement of collagen IV scaffolds with sulfilimine crosslinks may be substantially reduced (see supplemental methods). Indeed, smoking has been associated with architectural changes within BMs (Asmussen, 1979; Soltani et al., 2012). Finally, since BM assembly involves Br−, tissue development or remodeling may be vulnerable to Br-deficiency. The findings of our study provide rationale for investigating the clinical implications of Br-deficiency and the physiologic consequences of mechanically perturbing collagen IV scaffolds.

Experimental Procedures

Detailed Experimental Procedures for materials and methods appear in the Extended Experimental Methods.

High-Resolution Mass Spectrometry

NC1 domains were resolved by SDS-PAGE, excised, and “in-gel” trypsin digested as described (Vanacore et al., 2009). The resultant sample was enriched for sulfilimine crosslinked peptides, and analyzed using LC-MS. Data analysis used a combination Thermo Xcalibur 2.1, the Myrimatch algorithm with Bumbershoot suite, Scaffold, (Proteome Software, Portland, OR), ScanRanker, and IonMatcher software, where appropriate.

Br-free Salt Purification

Concentrated solutions of metal (Na, K) hydroxide and reagent-grade HCl were placed in a sealed chamber that prevented liquid mixing yet allowed vapor diffusion (Fig. 2C, Fig. S3). After 4 days, sufficient HCl vapor had had diffused and reacted to neutralization the OH−, and the resultant metal Cl− salt was assayed for purity via ICP-MS (Table S1).

PFHR-9 Cell Culture/collagen matrix preparation

PFHR-9 cells (ATCC® CRL-2423™) were cultured in Dulbecco’s Modified Eagle Medium (DMEM) with 5% FBS under 10% CO2. Collagen IV was deposited in culture for 7–9 days with daily media changes. Uncrosslinked hexamer was generated by culturing cells in either 50 µM phloroglucinol (PHG) or 1 mM KI. Br-free DMEM for cell culture was prepared as per Tables S3–4.

Protein Purification

Recombinant human peroxidasin (hPXDN) was expressed in HEK293 cells and purified as described previously (Bhave et al., 2012), with Br-free buffers used as appropriate in chromatography, dialysis, and centrifugation steps. Purified enzyme was reacted in vitro with uncrosslinked NC1 generated in PHG treated PHFR-9 culture. Uncrosslinked hexamers were isolated with collagenase digestion and purified by subsequent chromatography in normal or Br-free buffers (Extended Experimental Methods).

Chemical Crosslinking of NC1 domains by hypohalous acids

Uncrosslinked NC1 was reacted with hypohalous acid for 1 minute at 37°C, quenched with methionine, and analyzed by SDS-PAGE. HOBr was prepared as previously described (Bhave et al, 2012). Detailed methods for densiometric analysis and thermodynamic calculations are in the Extended Experimental Methods.

Br-free Drosophila Food

S. cerevisiae was cultured in an adapted Br-free yeast nitrogen base (YNB) media (Table S6). Phytagel-based fly media contained additional vitamins and minerals (Table S7) plus ampicillin and Tegocept. Br-free NaCl and KCl were the only sources of chloride in the fly media. The combination of Br-free yeast and phytagel was used for both larval and adult rearing. Final media Br− levels were undetectable by NAA (Table S2).

Drosophila Genetics and Methods

see Extended Experimental Methods for details.

Statistical Analysis

Analysis performed in this work was completed in GraphPad Prism v. 5.00 for Windows (GraphPad Software, San Diego California USA) and SPSS v. 22 (IBM). All statistical tests between groups were analyzed using non-parametric measures indicated unless the data was found to be normal.

Supplementary Material

01
02

Highlights.

  • Bromine is essential for basement membrane architecture and tissue development

  • Peroxidasin uses HOBr to form sulfilimine crosslinks in collagen IV scaffolds

  • Selectivity of a bromosulfonium intermediate is the basis for Br essentiality

  • Br depletion is lethal in Drosophila and phenocopies peroxidasin mutants

Acknowledgments

This work was supported by: NIH R01 DK18381 to B.G.H.; NIH F30 DK100094-01 to A.S.M.; NIGMS T32 GM07347 to the Vanderbilt Medical-Scientist Training Program, and the Canby Robinson Society in support of A.S.M.; NIH KO8 DK097306 and the Vanderbilt Physician Scientist Development Award to G.B, and NIH R01 GM073883 to A.P.M. We are grateful to K.L. Rose at Vanderbilt Mass Spectrometry Resource Center, J. Clanton, X. Wang, E. Shannon for assistance with Drosophila, and J. Williams and W.G. Jerome for assistance with Electron Microscopy and sample preparation, performed with use of the VUMC Cell Imaging Shared Resource, and to the Bloomington Drosophila Stock Center, Yale FlyTrap, and the VDRC for fly stocks. The technical assistance of P. Todd and M. Rafi is appreciated. We acknowledge S. Kraft for helpful chemistry discussions. CFC and BGH are co-founders of SULFILATEC, INC. CFC is a director and officer without salary of SULFILATEC, INC.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Author Contributions

All authors contributed as a team to experimental design, interpretation of data, and writing of manuscript. A.S.M. lead the effort on Br essentiality and chemical mechanisms. C.F.C lead the effort on the initial observations of Br function in collagen IV crosslinking. G.B designed the Br-free Cl purification method, all Br-free methods and their application to the overall project. A.S.M. and R.V. conducted mass spectrometry experiments. A.S.M., G.B., and A.P.M. designed and conducted Drosophila experiments.

References

  1. Anke M, Regius Á, Groppel B, Arnhold W. Essentiality of the trace element bromine. Acta Agronomica Hung. 1990;39:297–303. [Google Scholar]
  2. Armesto XL, Canle LM, Fernandez MI, Garca MV, Santaballa JA. First Steps in the Oxidation of Sulfur-Containing Amino Acids by Hypohalogenation: Very Fast Generation of Intermediate Sulfenyl Halides and Halosulfonium Cations. Tetrahedron. 2000;56:1103–1109. [Google Scholar]
  3. Asmussen I. Fetal cardiovascular system as influenced by maternal smoking. Clin Cardiol. 1979;2:246–256. doi: 10.1002/clc.4960020403. [DOI] [PubMed] [Google Scholar]
  4. Barratt TM, Walser M. Extracellular fluid in individual tissues and in whole animals: the distribution of radiosulfate and radiobromide. The Journal of Clinical Investigation. 1969;48:56–66. doi: 10.1172/JCI105974. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bhave G, Cummings CF, Vanacore RM, Kumagai-Cresse C, Ero-Tolliver IA, Rafi M, Kang JS, Pedchenko V, Fessler LI, Fessler JH, et al. Peroxidasin forms sulfilimine chemical bonds using hypohalous acids in tissue genesis. Nat Chem Biol. 2012;8:784–790. doi: 10.1038/nchembio.1038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Borchiellini C, Coulon J, Le Parco Y. The function of type IV collagen during Drosophila muscle development. Mech Dev. 1996;58:179–191. doi: 10.1016/s0925-4773(96)00574-6. [DOI] [PubMed] [Google Scholar]
  7. Bozeman PM, Learn DB, Thomas EL. Assay of the human leukocyte enzymes myeloperoxidase and eosinophil peroxidase. J Immunol Methods. 1990;126:125–133. doi: 10.1016/0022-1759(90)90020-v. [DOI] [PubMed] [Google Scholar]
  8. Brodie BB, Brand E, Leshin S. The use of bromide as a measure of extracellular fluid. J Biol Chem. 1939;130:555–563. [Google Scholar]
  9. Chmutova G, Ziganshina AY, Movchan A. Differences in the Electronic Structure of Sulfenyl Chlorides and Bromides as Possible Reasons for Their Different Reactivities toward Aromatic Compounds. Russ J Org Chem. 1999;35:56–61. [Google Scholar]
  10. Dahlstrom KA, Ament ME, Medhin MG, Meurling S. Serum trace elements in children receiving long-term parenteral nutrition. J Pediatr. 1986;109:625–630. doi: 10.1016/s0022-3476(86)80225-6. [DOI] [PubMed] [Google Scholar]
  11. Daley WP, Yamada KM. ECM-modulated cellular dynamics as a driving force for tissue morphogenesis. Curr Opin Genet Dev. 2013;23:408–414. doi: 10.1016/j.gde.2013.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fidler AL, Vanacore RM, Chetyrkin SV, Pedchenko VK, Bhave G, Yin VP, Stothers CL, Rose KL, McDonald WH, Clark TA, et al. A unique covalent bond in basement membrane is a primordial innovation for tissue evolution. Proc Natl Acad Sci U S A. 2014;111:331–336. doi: 10.1073/pnas.1318499111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Gotenstein JR, Swale RE, Fukuda T, Wu Z, Giurumescu CA, Goncharov A, Jin Y, Chisholm AD. The C. elegans peroxidasin PXN-2 is essential for embryonic morphogenesis and inhibits adult axon regeneration. Development. 2010;137:3603–3613. doi: 10.1242/dev.049189. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gould DB, Phalan FC, Breedveld GJ, van Mil SE, Smith RS, Schimenti JC, Aguglia U, van der Knaap MS, Heutink P, John SW. Mutations in Col4a1 cause perinatal cerebral hemorrhage and porencephaly. Science. 2005;308:1167–1171. doi: 10.1126/science.1109418. [DOI] [PubMed] [Google Scholar]
  15. Groos S, Reale E, Hunefeld G, Luciano L. Changes in epithelial cell turnover and extracellular matrix in human small intestine after TPN. J Surg Res. 2003;109:74–85. doi: 10.1016/s0022-4804(02)00094-x. [DOI] [PubMed] [Google Scholar]
  16. Gupta MC, Graham PL, Kramer JM. Characterization of alpha1(IV) collagen mutations in Caenorhabditis elegans and the effects of alpha1 and alpha2(IV) mutations on type IV collagen distribution. J Cell Biol. 1997;137:1185–1196. doi: 10.1083/jcb.137.5.1185. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Haigo SL, Bilder D. Global tissue revolutions in a morphogenetic movement controlling elongation. Science. 2011;331:1071–1074. doi: 10.1126/science.1199424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hudson BG, Tryggvason K, Sundaramoorthy M, Neilson EG. Alport's syndrome, Goodpasture's syndrome, and type IV collagen. N Engl J Med. 2003;348:2543–2556. doi: 10.1056/NEJMra022296. [DOI] [PubMed] [Google Scholar]
  19. Hynes RO. The extracellular matrix: not just pretty fibrils. Science. 2009;326:1216–1219. doi: 10.1126/science.1176009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Klopman G. Chemical reactivity J Am Chem Soc. 1968;90:223–234. [Google Scholar]
  21. Langeveld JP, Wieslander J, Timoneda J, McKinney P, Butkowski RJ, Wisdom BJ, Jr, Hudson BG. Structural heterogeneity of the noncollagenous domain of basement membrane collagen. J Biol Chem. 1988;263:10481–10488. [PubMed] [Google Scholar]
  22. Lin CQ, Bissell MJ. Multi-faceted regulation of cell differentiation by extracellular matrix. FASEB J. 1993;7:737–743. doi: 10.1096/fasebj.7.9.8330681. [DOI] [PubMed] [Google Scholar]
  23. Mayeno AN, Curran AJ, Roberts RL, Foote CS. Eosinophils preferentially use bromide to generate halogenating agents. J Biol Chem. 1989;264:5660–5668. [PubMed] [Google Scholar]
  24. Mertz W. The essential trace elements. Science. 1981;213:1332–1338. doi: 10.1126/science.7022654. [DOI] [PubMed] [Google Scholar]
  25. Miura Y, Nakai K, Suwabe A, Sera K. Trace elements in renal disease and hemodialysis. Nucl Instrum Meth B. 2002;189:443–449. [Google Scholar]
  26. Nagy P, Beal JL, Ashby MT. Thiocyanate is an efficient endogenous scavenger of the phagocytic killing agent hypobromous acid. Chem Res Toxicol. 2006;19:587–593. doi: 10.1021/tx050338c. [DOI] [PubMed] [Google Scholar]
  27. Nielsen FH. Ultratrace elements in nutrition: Current knowledge and speculation. J Trace Elem Exp Med. 1998;11:251–274. [Google Scholar]
  28. Oe P, Vis R, Meijer J, Van Langevelde F, Allon W, Meer Cvd VH. Bromine deficiency and insomnia in patients on dialysis. Trace Element Metabolism in Man and Animals, TEMA-4 Australian Academy of Science, Canberra. 1981:526–529. [Google Scholar]
  29. Olszowy HA, Rossiter J, Hegarty J, Geoghegan P, Haswell-Elkins M. Background Levels of Bromide in Human Blood. J Anal Toxicol. 1998;22:225–230. doi: 10.1093/jat/22.3.225. [DOI] [PubMed] [Google Scholar]
  30. Pastor-Pareja JC, Xu T. Shaping cells and organs in Drosophila by opposing roles of fat body-secreted Collagen IV and perlecan. Developmental Cell. 2011;21:245–256. doi: 10.1016/j.devcel.2011.06.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Pearson RG. Hard and soft acids and bases, HSAB, part II: Underlying theories. J Chem Educ. 1968;45:643. [Google Scholar]
  32. Peskin AV, Turner R, Maghzal GJ, Winterbourn CC, Kettle AJ. Oxidation of methionine to dehydromethionine by reactive halogen species generated by neutrophils. Biochemistry. 2009;48:10175–10182. doi: 10.1021/bi901266w. [DOI] [PubMed] [Google Scholar]
  33. Pavelka S, Babicky A, Vobecky M. Biological half-life of bromide in the rat depends primarily on the magnitude of sodium intake. Physiol Res. 2005;54:639–644. [PubMed] [Google Scholar]
  34. Piedade-Guerreiro J, Carmo Freitas M, Martinho E. Application of NAA in determining the inorganic contents of the fruit fly (Ceratitis capitata Wied.) and its artificial food. J Radioanal Nucl Chem. 1987;110:531–537. [Google Scholar]
  35. Poschl E, Schlotzer-Schrehardt U, Brachvogel B, Saito K, Ninomiya Y, Mayer U. Collagen IV is essential for basement membrane stability but dispensable for initiation of its assembly during early development. Development. 2004;131:1619–1628. doi: 10.1242/dev.01037. [DOI] [PubMed] [Google Scholar]
  36. Rodriguez A, Zhou Z, Tang ML, Meller S, Chen J, Bellen H, Kimbrell DA. Identification of immune system and response genes, and novel mutations causing melanotic tumor formation in Drosophila melanogaster. Genetics. 1996;143:929–940. doi: 10.1093/genetics/143.2.929. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Seeman JI. Effect of conformational change on reactivity in organic chemistry. Evaluations, applications, and extensions of Curtin-Hammett Winstein-Holness kinetics. Chem Rev. 1983;83:83–134. [Google Scholar]
  38. Shanbhag S, Tripathi S. Epithelial ultrastructure and cellular mechanisms of acid and base transport in the Drosophila midgut. J Exp Biol. 2009;212:1731–1744. doi: 10.1242/jeb.029306. [DOI] [PubMed] [Google Scholar]
  39. Soltani A, Muller HK, Sohal SS, Reid DW, Weston S, Wood-Baker R, Walters EH. Distinctive characteristics of bronchial reticular basement membrane and vessel remodelling in chronic obstructive pulmonary disease (COPD) and in asthma: they are not the same disease. Histopathology. 2012;60:964–970. doi: 10.1111/j.1365-2559.2011.04147.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Song JJ, Ott HC. Organ engineering based on decellularized matrix scaffolds. Trends Mol Med. 2011;17:424–432. doi: 10.1016/j.molmed.2011.03.005. [DOI] [PubMed] [Google Scholar]
  41. Trautner M, Wieth JO. Renal excretion of chloride, bromide and thiocyanate during water diuresis. Acta Physiol Scand. 1968;74:606–615. doi: 10.1111/j.1748-1716.1968.tb04271.x. [DOI] [PubMed] [Google Scholar]
  42. Tsuge K, Kataoka M, Seto Y. Cyanide and thiocyanate levels in blood and saliva of healthy adult volunteers. J Health Sci. 2000;46:343–350. [Google Scholar]
  43. van Leeuwen FX, Sangster B. The toxicology of bromide ion. Crit Rev Toxicol. 1987;18:189–213. doi: 10.3109/10408448709089861. [DOI] [PubMed] [Google Scholar]
  44. Van Logten MJ, Wolthuis M, Rauws AG, Kroes R, Den Tonkelaar EM, Berkvens H, Van Esch GJ. Semichronic toxicity study of sodium bromide in rats. Toxicology. 1974;2:257–267. doi: 10.1016/0300-483x(74)90017-1. [DOI] [PubMed] [Google Scholar]
  45. Vanacore R, Ham AJ, Voehler M, Sanders CR, Conrads TP, Veenstra TD, Sharpless KB, Dawson PE, Hudson BG. A sulfilimine bond identified in collagen IV. Science. 2009;325:1230–1234. doi: 10.1126/science.1176811. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Vracko R. Basal lamina scaffold-anatomy and significance for maintenance of orderly tissue structure. Am J Pathol. 1974;77:314–346. [PMC free article] [PubMed] [Google Scholar]
  47. Wallaeys B, Cornelis R, Mees L, Lameire N. Trace elements in serum, packed cells, and dialysate of CAPD patients. Kidney Int. 1986;30:599–604. doi: 10.1038/ki.1986.227. [DOI] [PubMed] [Google Scholar]
  48. Walser M, Rahill WJ. Renal tubular reabsorption of bromide compared with chloride. Clin Sci (Lond) 1966;30:191–205. [PubMed] [Google Scholar]
  49. Wang X, Harris RE, Bayston LJ, Ashe HL. Type IV collagens regulate BMP signalling in Drosophila. Nature. 2008;455:72–77. doi: 10.1038/nature07214. [DOI] [PubMed] [Google Scholar]
  50. Weiss SJ, Test ST, Eckmann CM, Roos D, Regiani S. Brominating oxidants generated by human eosinophils. Science. 1986;234:200–203. doi: 10.1126/science.3018933. [DOI] [PubMed] [Google Scholar]
  51. Werb Z. ECM and Cell Surface Proteolysis: Regulating Cellular Ecology. Cell. 1997;91:439–442. doi: 10.1016/s0092-8674(00)80429-8. [DOI] [PubMed] [Google Scholar]
  52. Wolf RL, Eadie GS. Reabsorption of bromide by the kidney. Am J Physiol. 1950;163:436–441. doi: 10.1152/ajplegacy.1950.163.2.436. [DOI] [PubMed] [Google Scholar]
  53. Woods AE, Carlton RF, Casto ME, Gleason GI. Environmental bromine in freshwater and freshwater organisms: factors affecting bioaccumulation. Bull Environ Contam Toxicol. 1979;23:179–185. doi: 10.1007/BF01769938. [DOI] [PubMed] [Google Scholar]
  54. Young PR, Hsieh L-S. General base catalysis and evidence for a sulfurane intermediate in the iodine oxidation of methionine. J Am Chem Soc. 1978;100:7121–7122. [Google Scholar]
  55. Yurchenco PD. Basement membranes: cell scaffoldings and signaling platforms. Cold Spring Harb Perspect Biol. 2011:3. doi: 10.1101/cshperspect.a004911. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS602466-supplement-01.pdf (7.5MB, pdf)
02
NIHMS602466-supplement-02.pdf (367KB, pdf)

RESOURCES